0
Select Articles

Fact Fest PUBLIC ACCESS

The Latest Methods In Data Acquisition Provide Engineers With Stronger Support Than Ever To Make More Informed Decisions.

[+] Author Notes

Associate Editor

Mechanical Engineering 122(08), 66-68 (Aug 01, 2000) (3 pages) doi:10.1115/1.2000-AUG-4

The latest methods in data acquisition provide engineers with stronger support than ever to make more informed decisions. Data acquisition is the process of making repetitive, timed measurements of multiple signal sources. The robotics team at the Oak Ridge National Laboratories builds a computerized model of robotic parts, such as robotic arm, and then runs a series of data acquisition tests on a prototype to determine how the arm will move. The robotic arm was constructed after engineers ran data acquisition tests on a prototype to learn how it would react to forces from the arm and from the human controlling the robot. Those results were considered when programming the robotic arm. Some companies use data acquisition methods to determine if it makes economic sense to implement a project. Data acquisition boards measure signals like friction, as measured here, and provide feedback that appears as numbers and wave measurements on the computer screen. Results can be downloaded to another software package, which displays the hard data as waves or graphs for analysis.

RESEARCHERS IN FIELDS ranging from mechanical engineering to the health sciences depend on raw data to learn about a process or structure they’re investigating. A doctor, for example, might be interested in measuring and studying brain activity to gain insight into reactions to a stimulus. A mechanical engineer may want a record of events as an engine’s torque level varies over time. The results may lead to a refined engine design for greater power or efficiency.

Through a method of acquiring raw data—aptly called data acquisition—both the brain and the engine are monitored and wired to computers. The feedback they give is measured as a series of signals that serve as a kind of map to those interpreting the data.

“Data acquisition is the process of making repetitive, timed measurements of multiple signal sources,” said Ed Carafa, president ofEJC Systems, a data acquisition consulting company and equipment supplier in Newfoundland, N.J.

“I think of data acquisition almost like photography,” Carafa added. “People take pictures for a variety of purposes. They document conditions, inspect objects, record events, or they take photos to do a detailed analysis. In the same way, data acquisition records sequential measurements that might be used to monitor a machine, test a product, or support research work.”

Lonnie Love, a control engineer at Oak Ridge National Laboratories in Oak Ridge, Tenn., for example, has long used data acquisition to glean important information on how a robot will act in certain situations. He uses the data to program the controllers that power the robots made by the laboratory. The laboratory, owned by the U.S. Department of Energy, mainly does energy research. The robotics division in which Love works primarily studies nuclear waste cleanup, he said.

“We look at how you get robots into places that are dangerous for humans to go and get those places cleaned up,” he said. The division also designs robots used by the U.S. military. Love programs the controllers that literally control the robot’s inner workings.

“The robot is just a bunch of motors and sensors,” Love said. “So we use data acquisition to figure out how we’re going to drive those motors, based on the information we get about them.”

One of the systems Love and his team recently designed is a robot that loads bombs on the wing of an airplane.

“It’s a big robot and it has to know how to interpret signals, both from the software that controls it and from the human giving signals,” Love said.

Force sensors on the robot tell it how much the bomb weighs so that it can exert the proper amount of force needed to lift the load. Other force sensors interpret the amount of force applied by the human giving the robot commands. The robot interprets all this information via the software program that runs the controller.

The robotics team at the Oak Ridge National Laboratories builds a computerized model of robotic parts, such as this robotic arm, then runs a series of data acquisition tests on a prototype to determine how the arm will move.

Grahic Jump LocationThe robotics team at the Oak Ridge National Laboratories builds a computerized model of robotic parts, such as this robotic arm, then runs a series of data acquisition tests on a prototype to determine how the arm will move.

The robot’s control program also must know how to process many other signals, such as position measurement signals, which tell the immediate configuration of the robot’s arm. Velocity signals give information on how fast the arm is moving as it carries bombs. All these signals have to be managed so that they, in turn, can be used to manage the motors on the robot. The software program Love and his colleagues wrote for the controller determines how much force and velocity the robot will have to apply in particular situations based on information from its force sensors.

But before Love and his colleagues could write the program, they had to understand how much force the robot needed to exert in many situations.

To determine these measures, Love and his team used what is known as a data acquisition board. The board sends signals to the robot and then gathers information on how the robot reacted to those signals.

The information from the board was then sent directly into a software program called Matlab, from Mathworks in Natick, Mass. The software, Love said, provided an easy way to measure the signals, to literally see them diagrammed as waves or as a type of graph that could be studied.

“It’s a powerful tool for looking at the amplitude or the frequency component of a signal and for visualizing it. It looks a lot like a graph,” Love said.

A new interface provided with the software let Love send signals directly through the data acquisition boards. He then saved the feedback information directly into the Matlab program for analysis. Prior to this interface, Love’s team members had to write software that communicated with the boards, and then they recorded the data feedback in a separate computer file. They had to load that file into the Matlab program for interpretation. Before the interface, Matlab couldn’t control the boards.

The robotic arm was constructed after engineers ran data acquisition tests on a prototype to learn how it would react to forces from the arm and from the human controlling the robot. Those results were taken into account when programming the robotic arm.

Grahic Jump LocationThe robotic arm was constructed after engineers ran data acquisition tests on a prototype to learn how it would react to forces from the arm and from the human controlling the robot. Those results were taken into account when programming the robotic arm.

In another project, Love’s group looked at how to build a robot that could do its job in an oven that reached temperatures up to 1,000°F. First, they built a new type of water-cooled robot. Then they used data acquisition methods to study how it worked in such high temperatures. Through the data acquisition process, they discovered that even while it labored in the 1,000° oven, the robot’s internal temperature never exceeded 140°F, which is well below the maximum point it could withstand and still function.

Researchers at Oak Ridge Labs use data acquisition in a number of ways that are not tied to the development of robots. They used it, for example, as part of an experiment that studied the ways an oil- cooled motor operated.

“We wanted to see the torque the motor was putting out, so we had a sensor to measure physical torque that we hooked up to the board,” Love said. “We had four thermocouples that provided different temperature measurements and those went into the boards. We had pressure measurements and fluid flow measures. We had eight or nine different signals sent to the boards to be recorded.”

Love’s team sent a varying current to the motor to study how those signals would be affected by varied motor torque. The signal information was then sent to the software for analysis.

“When it was done, I had a big buffer of data I could look at,” Love said. “I could look at the temperature and pressure variations to see how they affected torque.”

Data acquisition systems vary greatly, although their basic principle—the measurement of signals—remains the same, said Carafa ofEJC Systems. Some systems are capable of acquiring huge amounts of data, while others simply take a few measurements. Modern data acquisition systems can take millions of samples each second. They take measurements only milliseconds apart or take continuous readings, he said. But to some engineers, these measurement capabilities represent tremendous overkill. They might need only to record measurements on an hourly basis and they set up their systems accordingly.

Some companies use data acquisition methods to determine whether or not it makes economic sense to implement a project. For instance, one electric utility company turned to data acquisition because, as part of an effort to promote electricity, it wanted its meter readers and customer service employees to drive electric trucks on their rounds. But to make sure the idea was economically sound, engineers had to determine how long the battery could power the truck before it needed a recharge, according to Mike Sarnowski, a marketing director at Iotech, a Cleveland company that furnished a data acquisition system to the electric company.

To find how far the truck could go between charges, the company installed an electrical propulsion system in a small pickup truck. Then, the electrical engineers looked at ways to increase the truck’s electrical efficiency so it could travel farther between recharges.

They looked at four options: installing solar panels to extend battery life; installing regenerative brakes to recover kinetic energy that could be used to further power the truck; constructing the truck of lighter, plastic components; and installing low-resistance tires to cut friction.

Each of these methods would extend the vehicle s range between charges, but by how much? To learn how much extra mileage each change would mean for the electric trucks, engineers used data acquisition methodology to compare the distance the truck ran between recharges after each modification.

Because the truck had to be driven over diverse terrain to make sure the test mirrored real-world driving conditions, the researchers used a notebook PC, which could ride along with the driver and record truck conditions, including its speed.

To get started, the engineers connected the truck to sensors that measured voltage, current, and temperature. They used a current sensor to measure battery charge and discharge current. They also recorded temperatures in various locations on the truck to learn how heat affected battery charge. For instance, they measured the battery’s casing temperatures as well as the heat of the drive motor. They also recorded temperatures in the bed of the truck so they could compare the outside air temperature to the internal readings.

To determine how a machine will operate in real-life conditions, engineers connect a prototype with wires to a data acquisition system (not seen), which sends voltage signals to the machine and then measures its responses.

Grahic Jump LocationTo determine how a machine will operate in real-life conditions, engineers connect a prototype with wires to a data acquisition system (not seen), which sends voltage signals to the machine and then measures its responses.

The electric company is now mulling test results before making its final decision on outfitting a fleet of electric trucks. The company is also looking at different ways to use the data acquisition system, Sarnowski said. For example, it considered adding a global positioning satellite system to the electrical trucks. When drivers take trucks out on calls, they can use the system to find the vehicles’coordinates as well as the coordinates of their destinations. The on-board data-acquisition system would combine this information with terrain data and battery-discharge data, so drivers will know whether the trucks can make the trip or whether the batteries are likely to run down, leaving them standing on the side of the road.

Data acquisition boards measure signals like friction, as measured here, and provide feedback that appears as numbers and wave measurements on the computer screen. Results can be down loaded to another software package, which displays the hard data as waves or graphs for analysis.

Grahic Jump LocationData acquisition boards measure signals like friction, as measured here, and provide feedback that appears as numbers and wave measurements on the computer screen. Results can be down loaded to another software package, which displays the hard data as waves or graphs for analysis.

As with many other methods of gathering information, the use of the personal computer revolutionized the data acquisition field, Love said.

“It was very painful,” Love said of his early data acquisition days. “When I first started in engineering, we had a dynamic signal analyzer. That’s a big, fancy oscilloscope that lets you send out a signal and measure it simultaneously. The device cost $35,000, and it was a pain in the neck to figure out how to use.”

The engineers had to analyze the signal sent by the oscilloscope in order to find the information they sought.

“And now companies are taking tools that previously stood alone, like the oscilloscope, and they’re saying, We can make your computer behave like an oscilloscope,’ ” Love said.

Now an engineer uses a personal computer not only as a data acquisition tool, but also as a word processor, perhaps as a design station, and to communicate, by e-mail.

Today, data acquisition boards, the software programs that analyze data, and the interfaces to those programs can be easily purchased by companies large or small, and customized to fit their needs. This access makes data acquisition available to many companies that previously couldn’t afford to run these tests. And also makes it easy to use, Carafa said.

As these systems become easier to use, engineers and researchers are finding more and more ways to use them, he added. The use of data acquisition is expanding with the advent of easy-to-use software and notebook-size PCs that can be taken into the field. Carafa said he sees no end in sight to the ways researchers and engineers can use feedback they get from their data acquisition boards.

Copyright © 2000 by ASME
View article in PDF format.

References

Figures

Tables

Errata

Discussions

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In